Career Spotlight with high performers (Interview)
This week I’m the guest profile on Andrew Yeung’s Career Spotlight series. I share about my journey so far, how I get motivation to write a blog post a week, my life “philosophy” and more!
Andrew is a bizops lead at Facebook, and writes a newsletter, “Musings & Perspectives”, on growth, productivity, and performance.
Every month, Andrew features a high performer ‘Career Spotlight’ and interviews them on their journey, mindset, and habits. The following interview is cross-posted with Andrew’s site.
This month’s spotlight will be with Susan Shu Chang, Principal Data Scientist at Clearco and independent game developer.
Susan is currently a Principal Data Scientist at Clearco where she specializes in machine learning infrastructure for on-demand investment decision-making. 🤯
Susan and I both started our careers in telecommunications — working for one of Canada’s largest telecommunications companies. We both write a newsletter spanning the topics productivity, career development, and growth. And we’re both pretty active on Linkedin — always looking to meet folks to get new perspectives — which is how we connected.
Coming across her profile for the first time, I immediately saw how passionate Susan was toward the topics of self-development, growth, and giving back to the community. While doing a full time gig as a Principal Data Scientist, she also hosts YouTube livestreams on machine learning, builds video games as a hobby, takes part in public speaking engagements, and writes for her personal newsletter.
She’s deeply introspective, as you’ll see from her writing — breaking down what may seem like a surface level concept, through second and third-order thinking.
I’m incredibly grateful to have her here to share her thoughts on career, writing, and experimentation.
What do you currently do as Principal Data Scientist at Clearco? What does a Data Scientist do?
Hi! My name is Susan, and I am currently a Principal Data Scientist at Clearco.
Clearco, formerly known as Clearbanc, is an eCommerce investor providing non-dilutive growth capital to founders. Clearbanc’s co-founder and president, Michele Romanow, has been on the Dragon’s Den TV show (think Shark Tank in Canada), where she noticed that for early stage companies, raising funding via equity can be way too expensive for founders.
Clearco uses machine learning and automation to provide fast, non-dilutive funding, with less friction from human biases than traditional funding (banks, venture capital).
Naturally, the machine learning components are what my team works on. There are the models themselves, as well as the infrastructure that allows the inference of multiple models to be served on demand. I have been focusing on leading the ML infrastructure improvements as of late to focus on scale, as we have raised a Series C lately, putting Clearco at unicorn status.
To sum up: My team works on augmenting Clearco’s funding products with machine learning and data science. This includes ML model training, data quality, infrastructure, and automation.
You’ve got a ton of interesting things on your resume. What has your career journey been like? What do you do outside your full-time job?
My educational background was in Economics, which heavily focused on inferential and predictive modeling, using data from financial markets, product pricing, household earnings, and so on. Sounds quite a lot like data science in industry, doesn’t it?
During my studies, Econometrics gave me a solid understanding of statistics, and upper year courses required a lot of calculus and matrix algebra, which were invaluable when I started self-learning about machine learning algorithms.
But, there was a catch — I only learned statistical programming through a proprietary software called Stata, not the languages common in industry, such as Python or R.
However, I had been using Python for a couple of years due to programming video games for fun.
One day a friend mentioned to me, “you have these two skills, in statistics and programming — have you heard of a field called data science”?
I had not, and googled it that day. It was a perfect, almost accidental combination of my knowledge of statistics and Python. I write about this entire process in this two-part article series.
I still run my video game studio on weekends, outside of full time work, and I’m currently working on my second commercial (for sale) game. I am able to do this, thanks to my dedication to focus optimization and daily routines, which allows me time for these “extracurriculars”, along with spending time with people important to me!
Can you talk about game development? What games do you design and what motivates you to do so? What’s your end goal?
My end goal is to share unique worlds and experiences to others, as if they had lived those experiences themselves.
The reason why I enjoy working on games on the side, instead of full time, is due to the total creative control I can have. How else can I slip in references to Ariana Grande songs, or ML algorithms in a video game?
As to why software creation is enticing as a creator, in terms of video games, it is simply incredible for thousands of people to spend time with any creation of mine, feel emotions, and mull on it.
It’s like I can invite them into a piece of my imaginary world, through a keyboard and monitor and code.
As a creator, it is a uniquely exhilarating feeling that an abstraction in my mind can be materialized by typing on a keyboard, which can then make someone across the world feel wonder and joy, or even existentialist nightmares (quotes from player reviews).
I don’t quite have the physique, or crane operating skills, to build a physical structure like the next Taj Mahal. I believe that with software, we as builders have the power to incite feelings of awe and wonder just the same.
You write a personal blog on career advice, data science, productivity, and self-learning. After two years of blog writing, how have you optimized your process and what have you learned?
When I first started writing on my blog, a post took me weeks to complete. Since then, my speed has increased greatly, and for the last few months I have published posts on a weekly cadence.
I generally follow 3 steps:
- Creative flow, destructive flow
- Step 1 - Idea marination and bullet point outline
- Step 2 - Brain vomit draft (Creative flow)
- Step 3 - Edit and polish (Destructive flow)
Creative flow, destructive flow
I first came across this idea from a Scott Young blog post, where he elaborates: The “creative flow” is when one generates ideas, but with low standards and no editing, since “perfectionism pushes away from creating ideas.”
Then, after sufficient ideas or draft content have been created, one can go about polishing the ideas, which is the “destructive flow”.
Trying to do both at the same time is like trying to make a sculpture out of marble, but the block of marble keeps growing in size, covering the chisel work, causing there to be no progress.
Step 1 — Idea marination and bullet point outline
Each week, usually on Sunday morning, I pick a topic from a list of topics I’ve been gathering for months. If a good idea comes up that week or the morning of, I might write about it immediately instead of picking from the list.
Once the week’s topic is decided on, I use the morning to “marinate my thoughts”. I also refer to this process as “productive meditation”, which is detailed in this post. While doing dishes and making coffee, for example, I think about the structure of the blog post, and how to explain each point.
Once the points have been marinated, I return to my computer and jot down headers and sub-headers of the blog post. Nowadays I do this in Notion, but I did it on pen and paper for years. I recently changed the medium due to portability.
Step 2 — Brain vomit draft (Creative flow)
Internally I call this second step “brain vomit”, a disturbing yet accurate description. Remember how I mentioned that the creative flow and the destructive flow should be separate, to avoid conflicts?
Due to that reason, when I write my first draft, I actively avoid using backspace (apart from typos or obvious grammatical errors), and don’t worry about flow, sentence structure, or minor grammatical errors.
I follow the bullet point outline, and type out my stream of consciousness to explain each point.
I adhere to the pomodoro timer (50 minute writing, 10 minute break) throughout this process. Sometimes, the words don’t flow from my brain to the keyboard, but that’s fine.
When stuck, I sit in silence to think, instead of getting distracted and checking social media or email. The train of thought will come, it might just be in a tunnel right now. But if you get distracted, your brain won’t be ready to pick up on the thought once the train is back out of the tunnel.
Depending on the topic and how much I marinated the bullet points, writing the brain vomit draft can take 2 to 3 hours, or longer.
Step 3 — Edit and polish (Destructive flow)
During this step, I take time to polish the formatting of paragraphs and images, and look up hyperlinks for any sources and references. I then remove to-do notes and comments. Since my posts are formatted with Markdown, I also make sure there are no code errors (missing a [ for example).
Sometimes I end up rewriting or cutting entire sentences or paragraphs. Here I’m less adamant on separating the creative and destructive flows, since rewriting a paragraph is more like repainting a wall, rather than setting up the skeleton frame of a building from scratch.
Depending on how much I procrastinated in the morning and afternoon, I might edit an article more than once, making sure to take another break between editing rounds to freshen my mind.
I’ve also had quite a bit of experience writing fiction, the most recent being the 70,000 word story script I wrote for the video game I developed. I think this practice helped me a lot with blog writing, especially when it comes to managing writer’s block.
What is your life and career philosophy?
What motivates me every day is that there is so much to experience and learn in our lifetime.
I try to optimize for being able to explore these new experiences. In particular, one way of thinking, which is “8 years a lifetime” has helped me.
I want to call out a defining experience that inspired me to expand what I saw as “possible”. It was a talk by Vivienne Ming.
Now, she is an accomplished neuroscientist and leader in the AI field, and gives away patents to charitable organizations because she believes that it will help more people in need. She has contributed heavily to research in the intersection between behavioral economics and machine learning, and also founded multiple companies.
After hearing her talk, it made me think:
Many dreams that I had thought impossible, suddenly seemed closer.
I have not experienced the difficulties Vivienne Ming has — not even close.
The fact that she could come back from the brink of death and such a dark despair, and continue on to provide so much to science and humanity, showed me that there was more in me to give.
She also mentioned that she sees 8 years as a lifetime.
It was ambiguous in the way that I recall it, but I personally interpreted it was that we can feel free to, for example, become a chef and climb to expert status in 8 years or so. Once there, we are free to explore elsewhere, such as going from chef to engineer, instead of feeling locked down, simply because we are in a profession that was determined by a choice we made in adolescence (commonly, choosing a university major).
Optimize for exploration and fulfillment.
What unconventional advice have you received?
Perhaps this isn’t “unconventional”, but one thing that I’ve tried to instill in myself is to not get caught too much in spending habits.
Tech writer Eugene Yan has also mentioned this — practicing gratitude and avoiding being stuck on the treadmill of being dissatisfied with our current milestone, only to feel the same when we reach the next, “bigger and better” milestone.
So, it’s not only advice I’ve “received”, but have been practicing, for a long time. I think due to my upbringing I’ve actually rarely felt the need to consume a lot to be happy — instead, I like to create as entertainment.
Perhaps I am kind of weird in this regard… probably due to what I learned from sleeping on a yoga mat for 2 years.
After that experience, I really don’t think I care to impress other people with riches anymore. I earn more, to fuel what I create (hiring art contractors for video games isn’t cheap), and I let what I do and what I create speak for itself.
What advice would you give early career folks?
The first piece of advice here, is to experiment a bit, and wander around a bit.
There’s no 100% straight and direct way to do something, or become something.
By trying things out, you’ll be better at narrowing down what you like to do, and also what you’re good at.
If you’re interested in tech now, it’s definitely helpful to try out different flavors — perhaps you like web development more than data science or vice versa, and you might be able to figure that out before you’re 5 years into university, or a career.
Hackathons and side projects are great ways to do that type of experimentation.
The second piece of advice is that it’s almost never “too late” to make a different decision.
I started learning programming only in my 3rd year of university, and on top of that, I didn’t choose a university major that is a “straight line” into tech. If that wasn’t too late for me, it won’t be too late for you. And I know many people that pivoted to different fields way later after graduation.